Pre-registration

registration ordering unverified

first-spec commit 4c8ce8e · 2026-07-18T22:11:21Z

run generated · 2026-05-01T08:30:57Z

Run timestamp predates this path's first git-add commit (rebase, rename, or pre-git local run). Spec hash is still the path's first-add commit — not repository HEAD — but ordering is not a clean pre-registration proof.

Japan's 1990-2020 near-zero per-capita GDP growth regime (mean GDP/capita growth < 1%) coincided with a meaningful improvement in life expectancy at birth (+5 years), a sharp decline in average annual hours worked per employed person (~-15%), and a stable Cantril life-satisfaction trajectory at OECD-median levels. The pattern refutes the proposition that per-capita GDP growth is a binding constraint on retaining or improving baseline-wellbeing outcomes in advanced economies.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

SUPPORTED iff all four canonical metrics are met (premise + LE gain + hours decline + satisfaction stable). REFUTED if fewer than 2 are met OR if the GDP-growth premise fails (mean >= 1.5% would mean Japan is not a near-zero-growth case in the relevant window). PARTIAL if 2 or 3 are met. METHOD_VALID: requires WDI LE + WDI growth + PWT avh on disk for JPN 1990-2020; if PWT avh missing, downgrade to inconclusive on hours-worked metric. Cantril window is honest about its 2008-2020 coverage.

formal test & threshold

test:      japan_zero_growth_basic_wellbeing_quad_check_1990_2020
threshold: PRIMARY: 4-of-4 canonical metrics met.

Method

Template: multi_metric_checklist
Clustering: none
Sample: 1 countries · 1990 – 2020
Evidence type: canonical_case_multi_metric

Single-country canonical case, four-leg pattern-match. Distinct from the existing japan_stagnation_wellbeing_outcomes spec which tests a six-leg canonical wellbeing basket including suicide and fertility (degraded dimensions); this spec tests only the four legs the eco-socialist / degrowth literature centres -- LE, hours, satisfaction, employment -- and asserts that on those four the verdict is supported, while remaining silent on the additional dimensions covered in the companion spec. The two specs are designed to score the same Japan canonical case on different fairly-defined indicator baskets.

Data

Variable	Source	Transform
gdp_per_capita_growth outcome	world_bank_wdi:NY.GDP.PCAP.KD.ZGtier 2	level
life_expectancy_at_birth outcome	world_bank_wdi:SP.DYN.LE00.INtier 2	level
avg_annual_hours_worked_per_employed outcome	pwt:avhtier 3	level
cantril_life_satisfaction outcome	owid:happiness-cantril-laddertier 2	level
unemployment_rate outcome	world_bank_wdi:SL.UEM.TOTL.ZStier 2	level
japan_lost_decades_window treatment	constructed:indicator = 1 for JPN 1990-2020tier 5	indicator

● ready · ● pending · ● reconstruct-needed

Detailed result card

Result card — japan_zero_growth_basic_wellbeing_intact

Verdict: refuted

Reason: 1 metrics failed and 0 pending; cannot reach 4

Pre-registered rule: SUPPORT if >= 4 of 4 metrics met; REFUTE if <= 2 met (impossible to hit support).

Counts: 3 MET · 1 NOT_MET · 0 PENDING_DATA · 0 PENDING_EVAL

Primary country: JPN

Metric-by-metric

| # | Metric | Status | Observed | Threshold | Notes | |---|---|:---:|---:|---|---| | 1 | jpn_zero_growth_premise | MET | | mean(NY.GDP.PCAP.KD.ZG, 1990-2019) < 1.5% | JPN mean 1990-2019 = 0.961; threshold <1.5 | | 2 | jpn_life_expectancy_gain | MET | | LE_2019 - LE_1990 >= 4 years | JPN 2019-1990 diff = 5.519 (used 2019, 1990); threshold >=4 | | 3 | jpn_hours_worked_decline | MET | | avh_2019 / avh_1990 <= 0.90 | JPN 2019/1990 ratio = 0.834 (used 2019, 1990); threshold <=0.9 | | 4 | jpn_life_satisfaction_stable_oecd_median | NOT_MET | | max(|cantril_jpn_t - cantril_oecd_median_t|) <= 1.0 over 2008-2020 | max |JPN - OECD median| = 1.016 in 2019; threshold <=1 |

Claim

Japan's 1990-2020 near-zero per-capita GDP growth regime (mean GDP/capita growth < 1%) coincided with a meaningful improvement in life expectancy at birth (+5 years), a sharp decline in average annual hours worked per employed person (~-15%), and a stable Cantril life-satisfaction trajectory at OECD-median levels. The pattern refutes the proposition that per-capita GDP growth is a binding constraint on retaining or improving baseline-wellbeing outcomes in advanced economies.

Interpretation

The canonical-case pattern match is not satisfied: only 3 of 4 metrics met their thresholds, below the support threshold of 4. Note that for canonical-case hypotheses, a refutation can indicate either that the hypothesis is genuinely weak, that the metric set is mis-calibrated (too strict), or that the data substrate has systematic gaps. Review the PENDING_DATA / PENDING_EVAL metrics before accepting the refutation.

Steelman live concerns

See hypotheses/steelman/japan_zero_growth_basic_wellbeing_intact.md for the strongest opposing arguments. Canonical-case multi-metric evidence is a pattern match, not a causal identification — the result card should be read as 'outcome trajectory matches the predicted pattern to degree X' rather than 'policy P caused the outcome'.

Provenance

Vintages pinned in manifest.yaml. Full per-metric diagnostics in diagnostics.json. Machine-readable results in metric_results.parquet.

Notes

Companion to japan_stagnation_wellbeing_outcomes (v2). That spec uses a 6-indicator canonical wellbeing basket including suicide and fertility (degraded dims); this spec uses the 4-indicator degrowth-favourable basket. The two together form the full scoreboard. WDI + PWT inputs are on disk per repo notes. Cantril ladder via OWID coverage gap 1990-2007 is acknowledged.

In ordinary language

Results